An epitaxial control method for DFB laser chips

Abstract

The present application provides an epitaxial control method for DFB laser chips, which relates to the technical field of wafer epitaxial growth and includes: performing reverse simulation with the constraints of meeting expected epitaxial property characteristics and a predetermined epitaxial growth cycle to determine epitaxial growth control parameters; dividing and determining multiple growth intervals, setting growth indexes for the multiple growth intervals according to the expected epitaxial property characteristics to obtain multiple expected growth indexes; and performing epitaxial growth control according to the epitaxial growth control parameters, the multiple growth intervals, and the multiple expected growth indexes. Through the present application, the technical problem in the existing method that due to the inability to quickly and accurately adjust the growth control parameters according to the growth state of the wafer epitaxial layer during the wafer epitaxial growth process, the accuracy and uniformity of epitaxial growth control are poor, affecting the growth quality of the epitaxial layer can be solved, the growth quality of the epitaxial layer can be improved, and the stability and consistency of the performance of the DFB laser chip can be ensured.

Description

Technical Field

[0001] This application relates to the technical field of wafer epitaxial growth, and particularly to an epitaxial control method for DFB laser chips. Background Art

[0002] In the field of modern optical communication, DFB (Distributed Feedback) laser chips, as one of the key devices, are widely used in optical fiber communication systems. The performance of DFB lasers highly depends on the quality of their epitaxial layers. Especially during the chip manufacturing process, the accuracy and uniformity of epitaxial growth control are crucial. Characteristics such as the material composition, thickness, and crystal surface profile of the epitaxial layer directly determine key performance indicators of the laser, such as emission wavelength, output power, and modulation rate. Therefore, the control of the epitaxial growth process is essential for ensuring the performance of DFB laser chips.

[0003] Currently, epitaxial growth technology usually adopts Molecular Beam Epitaxy (MBE) or Metal-Organic Chemical Vapor Deposition (MOCVD) processes. In these processes, epitaxial growth control depends on preset growth parameters, such as temperature, gas flow rate, and growth rate. However, since the growth control parameters cannot be adaptively adjusted according to the monitored state of the epitaxial layer during the growth process, there are certain limitations in the existing epitaxial growth control methods.

[0004] In summary, in the existing epitaxial growth control methods for DFB laser chips, there are technical problems that due to the inability to quickly and accurately adjust the growth control parameters according to the growth state of the wafer epitaxial layer during the wafer epitaxial growth process, the accuracy and uniformity of epitaxial growth control are poor, affecting the growth quality of the epitaxial layer. Summary of the Invention

[0005] The purpose of this application is to provide an epitaxial control method for DFB laser chips to solve the technical problems in the existing epitaxial growth control methods for DFB laser chips that due to the inability to quickly and accurately adjust the growth control parameters according to the growth state of the wafer epitaxial layer during the wafer epitaxial growth process, the accuracy and uniformity of epitaxial growth control are poor, affecting the growth quality of the epitaxial layer.

[0006] In view of the above problems, the present application provides an epitaxial control method for a DFB laser chip, including: reading the wafer attribute information of the wafer to be processed and the expected performance indicators of the DFB laser chip, aiming to meet the expected performance indicators, and analyzing and obtaining the expected epitaxial attribute characteristics of the epitaxial layer according to the wafer attribute information; obtaining a predetermined epitaxial growth cycle, using the TACD simulation tool for reverse simulation with meeting the expected epitaxial attribute characteristics and the predetermined epitaxial growth cycle as constraints to determine the epitaxial growth control parameters; dividing and determining a plurality of growth intervals based on the predetermined epitaxial growth cycle, setting growth indicators for the plurality of growth intervals according to the expected epitaxial attribute characteristics to obtain a plurality of expected growth indicators; and performing epitaxial growth control on the wafer to be processed according to the epitaxial growth control parameters, the plurality of growth intervals, and the plurality of expected growth indicators.

[0007] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0008] By reading the wafer attribute information of the wafer to be processed and the expected performance indicators of the DFB laser chip, aiming to meet the expected performance indicators, the expected epitaxial property characteristics of the epitaxial layer are analyzed and obtained according to the wafer attribute information; then a predetermined epitaxial growth cycle is obtained. With the constraints of meeting the expected epitaxial property characteristics and the predetermined epitaxial growth cycle, reverse simulation is carried out using the TACD simulation tool to determine the epitaxial growth control parameters; on the other hand, multiple growth intervals are determined based on the predetermined epitaxial growth cycle, and growth indicators are set for the multiple growth intervals according to the expected epitaxial property characteristics to obtain multiple expected growth indicators; finally, the epitaxial growth control of the wafer to be processed is performed according to the epitaxial growth control parameters, multiple growth intervals and multiple expected growth indicators; wherein the method for performing epitaxial growth control includes: S1: Select the first growth interval and the second growth interval, and the second expected growth indicator of the second growth interval, wherein the first growth interval is the first growth interval to start among the multiple growth intervals, and the second growth interval is the adjacent growth interval of the first growth interval; S2: According to the epitaxial growth control parameters, perform epitaxial growth of the first growth interval on the wafer to be processed, wherein the epitaxial growth control parameters include growth temperature, reaction gas flow rate and reaction gas ratio; S3: Use reflection high-energy electron diffraction spot monitoring to obtain a crystal growth data set, wherein the crystal growth data set includes an epitaxial layer thickness set and a crystal surface profile set; S4: Perform crystal growth prediction within a predetermined window according to the epitaxial layer thickness set and the crystal surface profile set to obtain first predicted epitaxial property data at the end node of the first growth interval; S5: Calculate the deviation between the second expected growth indicator and the first predicted epitaxial property data to determine the second growth deviation; S6: With the constraints of meeting the second growth deviation and the second growth interval, perform reverse simulation using the TACD simulation tool to determine the second epitaxial growth control parameters; S7: Perform epitaxial growth control of the second growth interval according to the second epitaxial growth control parameters; Iteratively loop through S1 to S7 until the predetermined epitaxial growth cycle is completed. Through the above method, the technical goal of quickly and accurately adjusting the growth control parameters according to the growth state of the wafer epitaxial layer can be achieved, effectively solving the problems of insufficient accuracy and uniformity of epitaxial growth, thereby improving the growth quality of the epitaxial layer and ensuring the stability and consistency of the performance of the DFB laser chip.

[0009] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the following specifically illustrates the specific embodiments of the present application. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become easily understood through the following description. Brief Description of the Drawings

[0010] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings described below are merely exemplary. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on the provided drawings.

[0011] Figure 1 It is a schematic flowchart of an epitaxial control method for a DFB laser chip according to the present application;

[0012] Figure 2 It is a schematic flowchart of performing epitaxial growth control on a wafer to be processed in an epitaxial control method for a DFB laser chip according to the present application. Detailed Embodiments

[0013] By providing an epitaxial control method for a DFB laser chip, the present application solves the technical problem in the existing epitaxial growth control method for a DFB laser chip that due to the inability to quickly and accurately adjust the growth control parameters according to the growth state of the wafer epitaxial layer during the wafer epitaxial growth process, the accuracy and uniformity of the epitaxial growth control are poor, affecting the growth quality of the epitaxial layer. It can achieve the technical goal of quickly and accurately adjusting the growth control parameters according to the growth state of the wafer epitaxial layer, effectively solve the problem of insufficient accuracy and uniformity of the epitaxial growth, thereby improving the growth quality of the epitaxial layer and ensuring the stability and consistency of the performance of the DFB laser chip.

[0014] Next, the technical solutions in the present application will be clearly and completely described with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the example embodiments described here. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application. Additionally, it should be noted that for the sake of convenience of description, only the parts related to the present application are shown in the drawings rather than all of them.

[0015] Embodiment. Please refer to the attached Figure 1 , the present application provides an epitaxial control method for a DFB laser chip, which specifically includes the following steps:

[0016] Step 1: Read the wafer attribute information of the wafer to be processed and the expected performance indicators of the DFB laser chip. For the purpose of meeting the expected performance indicators, analyze and obtain the expected epitaxial attribute characteristics of the epitaxial layer according to the wafer attribute information.

[0017] Specifically, in the manufacturing process of DFB laser chips, the epitaxial growth quality is crucial for the performance of the chips. To ensure that the chips meet the expected performance indicators, it is necessary to control the epitaxial growth process based on the wafer attribute information and the expected performance indicators of the chips. First, read the wafer attribute information of the wafer to be processed. The wafer attribute information directly affects the growth process and quality of the epitaxial layer. The wafer attribute information includes wafer material, wafer lattice constant, surface roughness, and wafer thickness. On the other hand, read the expected performance indicators of the DFB laser chip. The expected performance indicators include wavelength, power, and modulation rate, which can be set according to the actual application scenarios and requirements of the chip.

[0018] Next, with the purpose of meeting the expected performance indicators for wafer epitaxial growth, conduct an analysis of the epitaxial attribute characteristics of the epitaxial layer according to the wafer attribute information. An attribute analysis model can be constructed based on a BP neural network, and sample data can be collected to train the attribute analysis model until it converges. Analyze the epitaxial attribute characteristics of the expected performance indicators to obtain the expected epitaxial attribute characteristics. The expected epitaxial attribute characteristics include expected material composition, expected epitaxial layer thickness, and expected crystal surface profile. By obtaining the expected epitaxial attribute characteristics of the epitaxial layer, it provides a basis for guiding the adjustment of growth control parameters during the epitaxial growth process, ensuring that the epitaxial growth quality of the final DFB laser chip meets the expected performance requirements.

[0019] Step 2: Obtain a predetermined epitaxial growth cycle. With the constraints of meeting the expected epitaxial attribute characteristics and the predetermined epitaxial growth cycle, use the TACD simulation tool for reverse simulation to determine the epitaxial growth control parameters.

[0020] Specifically, obtain a predetermined epitaxial growth period, where the epitaxial growth period is the entire process from the start to the end of epitaxial growth and the set expected growth time, which can be set by those skilled in the art according to the actual situation. For example, the expected growth time is set to 12 hours. Then, with the constraints of meeting the expected epitaxial property characteristics and the predetermined epitaxial growth period, use the TACD simulation tool for reverse simulation. First, select a TCAD simulation tool suitable for the epitaxial growth process, such as Synopsys Sentaurus Process. Such tools can simulate physical properties such as material deposition, growth rate, and surface topography. Then, input the expected epitaxial property characteristics (material composition, epitaxial layer thickness, and crystal surface profile) and the predetermined epitaxial growth period into the TCAD simulation tool, and automatically adjust the epitaxial growth control parameters (growth temperature, reaction gas flow rate, and reaction gas ratio). Through iterative simulation, continuously adjust the growth control parameters until the thickness, material composition, surface topography, etc. of the epitaxial growth reach the expected target values, and output the epitaxial growth control parameters that meet the expected epitaxial property characteristics and the predetermined epitaxial growth period. The epitaxial growth control parameters include growth temperature, reaction gas flow rate, and reaction gas ratio. By obtaining the epitaxial growth control parameters, it provides a basis for subsequent epitaxial growth control of the wafer.

[0021] Step 3: Based on the determined division of the predetermined epitaxial growth period, determine multiple growth intervals, and set growth indicators for the multiple growth intervals according to the expected epitaxial property characteristics to obtain multiple expected growth indicators.

[0022] Specifically, first, divide the predetermined epitaxial growth period into multiple growth intervals according to a predetermined time step. The predetermined time step can be set by oneself according to the actual situation. For example, set the predetermined time step to 30 minutes to obtain multiple growth intervals, where the multiple growth intervals have sequential identifiers. Then, set growth indicators for the multiple growth intervals according to the expected epitaxial property characteristics, that is, configure the expected growth indicators for the end nodes of each growth interval, including the material composition, epitaxial layer thickness, and crystal surface profile of each end node. By configuring the expected growth indicators for the end nodes of each growth interval, the fineness of wafer epitaxial growth control can be improved, ensuring precise control of the epitaxial growth process to achieve a high-quality epitaxial layer structure of the final DFB laser chip.

[0023] Step 4: Execute the epitaxial growth control of the wafer to be processed according to the epitaxial growth control parameters, multiple growth intervals, and multiple expected growth indicators.

[0024] Specifically, finally, perform epitaxial growth control on the wafer to be processed according to the epitaxial growth control parameters, multiple growth intervals, and multiple expected growth indicators, that is, in the order of the growth intervals, perform real-time monitoring and feedback adjustment on the wafer epitaxial growth according to multiple expected growth indicators to ensure that the growth quality at any stage during the epitaxial growth process meets the expected requirements, and finally achieve high-quality epitaxial layer growth to meet the performance indicators of the DFB laser chip.

[0025] By providing an epitaxial control method for a DFB laser chip, it is possible to solve the technical problem in the existing epitaxial growth control method for a DFB laser chip that due to the inability to quickly and accurately adjust the growth control parameters according to the growth state of the wafer epitaxial layer during the wafer epitaxial growth process, the accuracy and uniformity of the epitaxial growth control are poor, affecting the growth quality of the epitaxial layer. By dividing a predetermined epitaxial growth cycle into multiple growth intervals, setting growth indicators for the multiple growth intervals according to the expected epitaxial property characteristics, and obtaining multiple expected growth indicators; then selecting a first growth interval and a second growth interval, as well as the second expected growth indicator of the second growth interval; according to the epitaxial growth control parameters, perform epitaxial growth of the first growth interval on the wafer to be processed, and obtain a crystal growth data set by using reflection high-energy electron diffraction spot monitoring; further perform crystal growth prediction within a predetermined window according to the epitaxial layer thickness set and the crystal surface profile set to obtain the first predicted epitaxial property data at the end node of the first growth interval; then calculate the deviation between the second expected growth indicator and the first predicted epitaxial property data to determine the second growth deviation; and use the TACD simulation tool for reverse simulation with satisfying the second growth deviation and the second growth interval as constraints to determine the second epitaxial growth control parameter; finally, perform epitaxial growth control of the second growth interval according to the second epitaxial growth control parameter, and iterate the above steps until the predetermined epitaxial growth cycle is completed; it is possible to achieve the technical goal of quickly and accurately adjusting the growth control parameters according to the growth state of the wafer epitaxial layer, effectively solve the problem of insufficient accuracy and uniformity of epitaxial growth, thereby improving the growth quality of the epitaxial layer and ensuring the stability and consistency of the performance of the DFB laser chip.

[0026] Further, this application includes:

[0027] The wafer attribute information includes wafer material, wafer lattice constant, surface roughness, and wafer thickness, and the expected performance indicators include wavelength, power, and modulation rate.

[0028] Specifically, the wafer attribute information includes wafer material, wafer lattice constant, surface roughness, and wafer thickness. Among them, the wafer material is the basis for the epitaxial layer to be processed, directly affecting the lattice matching and growth behavior of the epitaxial layer; the wafer lattice constant is the spatial periodic constant of the atomic arrangement in the crystal structure. When the epitaxial layer grows, the lattice constant of the epitaxial material should be as close as possible to that of the substrate wafer to reduce the stress and defects caused by lattice mismatch; the surface roughness refers to the flatness of the wafer surface; the wafer thickness is the physical thickness of the wafer material, which affects the heat conduction and mechanical strength of epitaxial growth.

[0029] The expected performance indicators include wavelength, power, and modulation rate. Among them, the wavelength is a characteristic parameter of the light emitted by the DFB laser and is crucial for meeting communication standards. The wavelength of the laser is determined by the material composition and thickness of the epitaxial layer; the power is the output energy of the laser. The power of the laser is determined by the material quality, thickness, and electrical characteristics of the epitaxial layer. A high-quality epitaxial layer can improve the optoelectronic conversion efficiency of the laser and thus increase the output power; the modulation rate is the rate at which the laser modulates the optical signal, usually measured in GHz (gigahertz), referring to the speed at which the laser can perform signal modulation. The modulation rate is closely related to the electrical characteristics of the epitaxial layer (such as carrier mobility, injection efficiency, etc.), especially related to the epitaxial thickness of the active region.

[0030] Furthermore, this application includes:

[0031] Taking the wafer attribute information as a constraint and wafer epitaxial growth as a guide, retrieve and obtain a sample data set. Among them, the sample data includes sample performance indicators and sample epitaxial attribute features, and the sample epitaxial attribute features include material composition, epitaxial layer thickness, and crystal surface profile; use the sample data set to perform supervised training on the BP neural network to obtain an attribute analyzer; through the attribute analyzer, analyze the expected performance indicators and output the expected epitaxial attribute features.

[0032] Specifically, first, take the wafer attribute information and the expected epitaxial growth process as retrieval constraint conditions. Among them, the expected epitaxial growth process can be selected according to the actual situation, including molecular beam epitaxy (MBE), metalorganic chemical vapor deposition (MOCVD) process, etc. Take wafer epitaxial growth as the retrieval guiding word and retrieve sample data based on big data technology to obtain a sample data set. Among them, the sample data includes sample performance indicators and sample epitaxial attribute features, and the sample epitaxial attribute features include material composition, epitaxial layer thickness, and crystal surface profile.

[0033] Next, an initial attribute analyzer is constructed based on the BP neural network. The initial attribute analyzer is a neural network model in machine learning that can be iteratively optimized, including an input layer, multiple hidden layers, and an output layer. The input data of the input layer is performance indicators, and the output data is the extrinsic attribute features. Then, the sample data set is divided into a training set and a validation set according to a predetermined ratio. Usually, the training set accounts for 80% and the validation set accounts for 20%. Further, the initial attribute analyzer is supervised and trained using the training set, that is, supervised learning is performed using the loss function and the gradient descent algorithm to optimize and correct the weights and biases in the model, and an initial attribute analyzer in a convergent state is obtained. Then, the initial attribute analyzer is verified and trained using the validation set until an attribute analyzer that meets the expected conditions is obtained. By constructing an attribute analyzer based on the BP neural network for extrinsic attribute feature analysis, the efficiency and accuracy of obtaining extrinsic attribute features can be improved.

[0034] Finally, the expected performance indicators are input into the attribute analyzer for parsing, and the expected extrinsic attribute features are output.

[0035] Further, as shown in the appendix Figure 2 This application includes:

[0036] S1: Select a first growth interval and a second growth interval, and a second expected growth index of the second growth interval. Among them, the first growth interval is the first growth interval to start among the multiple growth intervals, and the second growth interval is an adjacent growth interval of the first growth interval; S2: According to the epitaxial growth control parameters, perform epitaxial growth of the first growth interval on the wafer to be processed, where the epitaxial growth control parameters include growth temperature, reaction gas flow rate, and reaction gas ratio; S3: Use reflection high-energy electron diffraction fixed-point monitoring to obtain a crystal growth data set, where the crystal growth data set includes an epitaxial layer thickness set and a crystal surface profile set; S4: Perform crystal growth prediction within a predetermined window according to the epitaxial layer thickness set and the crystal surface profile set to obtain first predicted extrinsic attribute data at the end node of the first growth interval; S5: Calculate the deviation between the second expected growth index and the first predicted extrinsic attribute data to determine the second growth deviation; S6: Using the TACD simulation tool for reverse simulation with the second growth deviation and the second growth interval as constraints to determine the second epitaxial growth control parameters; S7: Perform epitaxial growth control of the second growth interval according to the second epitaxial growth control parameters.

[0037] Specifically, first, select a first growth interval and a second growth interval from the multiple growth intervals, where the first growth interval is the earliest-starting growth interval among the multiple growth intervals, and the second growth interval is an adjacent growth interval to the first growth interval, and obtain a second expected growth index of the second growth interval. Then, according to the epitaxial growth control parameters, perform epitaxial growth of the first growth interval on the wafer to be processed, where the epitaxial growth control parameters include growth temperature, reaction gas flow rate, and reaction gas ratio; configure predetermined monitoring nodes, where the time intervals between adjacent monitoring nodes are the same and can be set according to requirements, such as monitoring once every 1 minute, and during the epitaxial growth of the first growth interval, use reflection high-energy electron diffraction (RHEED) to perform fixed-point monitoring of the wafer at the predetermined monitoring nodes. For example, through the periodic change of the RHEED stripes, the thickness of the epitaxial layer per minute is monitored in real time. The change period of the stripes is inversely proportional to the thickness of the epitaxial layer. By analyzing the stripe period, the grown thickness is calculated; the intensity and clarity of the RHEED pattern reflect the surface roughness and the step structure of the crystal. By analyzing these patterns, the change of the surface roughness is recorded; obtain a crystal growth data set, where the crystal growth data set includes an epitaxial layer thickness set and a crystal surface profile set.

[0038] Further, perform crystal growth prediction within a predetermined window according to the epitaxial layer thickness set and the crystal surface profile set, where the predetermined window is the sum of the monitoring data delay, the prediction data delay, and the parameter control delay, that is, to solve the lag of the monitoring data, predict the wafer growth state at the end node of the first growth interval according to the monitoring data, and obtain the first predicted epitaxial attribute data at the end node of the first growth interval. Then calculate the deviation between the second expected growth index and the first predicted epitaxial attribute data to obtain a second growth deviation, and the second growth deviation includes material component deviation, epitaxial layer thickness deviation, and crystal surface profile deviation.

[0039] Next, with the second growth deviation and the second growth interval as constraints, use the TACD simulation tool to perform reverse simulation, output the second epitaxial growth control parameters, and perform epitaxial growth control of the second growth interval according to the second epitaxial growth control parameters, which can improve the fineness of the wafer epitaxial growth control and achieve the technical goal of quickly and accurately adjusting the growth control parameters according to the growth state of the wafer epitaxial layer.

[0040] Furthermore, this application includes:

[0041] After performing epitaxial growth control of the second growth interval according to the second epitaxial growth control parameters, iterate through S1 to S7 until the predetermined epitaxial growth cycle is completed.

[0042] Specifically, after performing the epitaxial growth control of the second growth interval according to the second epitaxial growth control parameter, in the same way, iterate through S1 to S7 in a loop to continue optimizing and adjusting the epitaxial growth control parameters of the third growth interval, the fourth growth interval, and subsequent growth intervals until the predetermined epitaxial growth cycle is completed. By performing iterative feedback control, the quality and performance of the epitaxial layer can be ensured to be stable throughout the growth cycle and finally meet the predetermined design requirements.

[0043] Furthermore, this application includes:

[0044] Perform thickness growth trend analysis based on the epitaxial layer thickness set to determine the thickness growth rate; select the end epitaxial layer thickness in the epitaxial layer thickness set, and perform crystal growth prediction within a predetermined window based on the thickness growth rate and the end epitaxial layer thickness to obtain the first predicted epitaxial layer thickness; perform contour change trend analysis based on the crystal surface contour set to determine the contour change characteristics; select the end surface contour in the crystal surface contour set, and perform crystal growth prediction within a predetermined window based on the contour change characteristics and the end surface contour to obtain the first predicted crystal surface contour; obtain the first predicted epitaxial property data based on the first predicted epitaxial layer thickness and the first predicted crystal surface contour.

[0045] Specifically, first, perform thickness growth trend analysis based on the epitaxial layer thickness set. By plotting a thickness growth trend graph with the horizontal axis representing time (in minutes) and the vertical axis representing the thickness of the epitaxial layer (in nanometers), it can be seen from the trend graph whether the epitaxial growth is uniform, or whether there are fluctuations or anomalies; then calculate the thickness growth rate, which is the increase in thickness per unit time and can be obtained by averaging the overall growth rate of the entire epitaxial growth process in the first growth interval. Then select the end epitaxial layer thickness in the epitaxial layer thickness set, and the end epitaxial layer thickness is the epitaxial layer thickness at the last monitoring node; further perform crystal growth prediction within a predetermined window based on the thickness growth rate and the end epitaxial layer thickness, that is, calculate the predicted epitaxial layer thickness at the end node of the first growth interval to obtain the first predicted epitaxial layer thickness.

[0046] On the other hand, by performing contour change trend analysis based on the set of crystal surface contours, it is possible to draw a trend graph of the change in surface roughness over time and a graph of the change in fringe intensity. Through these graphs, the change pattern of the surface contour over time can be observed to determine the contour change characteristics, where the contour change characteristics include the changes in surface roughness, fringe intensity, and step structure. Then, select the end surface contour from the set of crystal surface contours, where the end surface contour is the wafer surface contour of the last monitoring node. Further, based on the contour change characteristics and the end surface contour, predict the crystal growth within a predetermined window to obtain the first predicted crystal surface contour. Finally, analyze the first predicted epitaxial layer thickness and the first predicted crystal surface contour to obtain the first predicted material composition, and set the first predicted epitaxial layer thickness, the first predicted crystal surface contour, and the first predicted material composition as the first predicted epitaxial property data.

[0047] Furthermore, this application includes:

[0048] Take the sum of the monitoring data delay, the prediction data delay, and the parameter control delay as the predetermined window.

[0049] Specifically, during the epitaxial growth control process, monitoring data, prediction data, and parameter control operations may incur different delays. To more accurately predict crystal growth, the sum of the monitoring data delay, the prediction data delay, and the parameter control delay can be used as the predetermined window, that is, the total duration of these delay factors is considered to set the prediction window in order to make timely and accurate control adjustments. The monitoring data delay refers to the delay that occurs when data collected during real-time monitoring using RHEED or other equipment during epitaxial growth is recorded and transmitted to the control system; the prediction data delay refers to the delay that may occur during the calculation and model processing when predicting the crystal growth trend based on existing monitoring data; the parameter control delay refers to the time required for adjusting the growth parameters to have an actual impact on the epitaxial layer growth during the process of controlling epitaxial growth. By setting a predetermined window for prediction, appropriate adjustments can be ensured during the period when the delay takes effect, improving the accuracy and real-time performance of epitaxial growth control, and ultimately contributing to the realization of high-quality epitaxial layer growth.

[0050] Furthermore, this application includes:

[0051] Using the TACD simulation tool, perform growth simulation on the wafer to be processed according to the first growth interval and epitaxial growth control parameters, and obtain the first simulated epitaxial layer thickness, the first simulated crystal surface profile, and the first simulated material composition; calculate the deviation between the first predicted epitaxial layer thickness and the first simulated epitaxial layer thickness to obtain the first predicted thickness difference; perform deviation analysis on the first predicted crystal surface profile and the first simulated crystal surface profile to determine the first predicted profile deviation characteristics; perform correlation correction on the first simulated material composition according to the first predicted thickness difference and the first predicted profile deviation characteristics to obtain the first predicted material composition; set the first predicted epitaxial layer thickness, the first predicted crystal surface profile, and the first predicted material composition as the first predicted epitaxial property data.

[0052] Specifically, first, use the TACD simulation tool to perform growth simulation on the wafer to be processed according to the first growth interval and epitaxial growth control parameters, and obtain the first simulated epitaxial layer thickness, the first simulated crystal surface profile, and the first simulated material composition. Then calculate the deviation between the first predicted epitaxial layer thickness and the first simulated epitaxial layer thickness to obtain the first predicted thickness difference; perform deviation analysis on the first predicted crystal surface profile and the first simulated crystal surface profile to determine the first predicted profile deviation characteristics, where the predicted profile deviation characteristics include surface roughness deviation, fringe intensity deviation, and step structure deviation.

[0053] Then perform correlation correction on the first simulated material composition according to the first predicted thickness difference and the first predicted profile deviation characteristics. A correction model can be constructed based on historical data analysis. For example, the expression of the correction model is ; where is the corrected i-th material composition, is the simulated material composition of the i-th material, is the thickness correction coefficient, is the surface profile correction coefficient, is the predicted thickness difference, is the predicted profile deviation characteristic. Using the correction model, correct the first simulated material composition according to the thickness difference and the surface profile deviation. By adjusting the proportion of each component, obtain the first predicted material composition. For example, if the In content is 50% in the simulation, but after correction due to thickness deviation and surface roughness, the predicted In content is 48%; obtain the first predicted material composition. Finally, set the first predicted epitaxial layer thickness, the first predicted crystal surface profile, and the first predicted material composition as the first predicted epitaxial property data.

[0054] In summary, the epitaxial control method for a DFB laser chip provided by this application has the following technical effects:

[0055] By dividing a predetermined epitaxial growth cycle into multiple growth intervals, setting growth indicators for the multiple growth intervals according to expected epitaxial property characteristics to obtain multiple expected growth indicators; then selecting a first growth interval and a second growth interval, as well as a second expected growth indicator for the second growth interval; according to the epitaxial growth control parameters, performing epitaxial growth of the first growth interval on the wafer to be processed, and obtaining a crystal growth data set by using reflection high energy electron diffraction spot monitoring; further predicting crystal growth within a predetermined window according to the epitaxial layer thickness set and the crystal surface profile set to obtain first predicted epitaxial property data at the end node of the first growth interval; then calculating the deviation between the second expected growth indicator and the first predicted epitaxial property data to determine the second growth deviation; and using the TACD simulation tool to perform reverse simulation with satisfying the second growth deviation and the second growth interval as constraints to determine the second epitaxial growth control parameters; finally, performing epitaxial growth control of the second growth interval according to the second epitaxial growth control parameters, and iteratively repeating the above steps until the predetermined epitaxial growth cycle is completed; it can achieve the technical goal of quickly and accurately adjusting the growth control parameters according to the growth state of the wafer epitaxial layer, effectively solve the problems of insufficient accuracy and uniformity of epitaxial growth, thereby improving the growth quality of the epitaxial layer and ensuring the stability and consistency of the performance of the DFB laser chip.

[0056] The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0057] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is also intended to include these changes and modifications.

Claims

1. An epitaxial control method for a DFB laser chip, characterized in that Including: Read the wafer attribute information of the wafer to be processed and the expected performance indicators of the DFB laser chip. For the purpose of meeting the expected performance indicators, analyze and obtain the expected epitaxial property characteristics of the epitaxial layer according to the wafer attribute information. Obtain a predetermined epitaxial growth cycle. With the expected epitaxial property characteristics and the predetermined epitaxial growth cycle as constraints, perform reverse simulation using the TACD simulation tool to determine the epitaxial growth control parameters. Based on the determined predetermined epitaxial growth cycle, divide it into multiple growth intervals. Set growth indicators for the multiple growth intervals according to the expected epitaxial property characteristics to obtain multiple expected growth indicators. Execute the epitaxial growth control of the wafer to be processed according to the epitaxial growth control parameters, multiple growth intervals, and multiple expected growth indicators. Among them, the wafer attribute information includes wafer material, wafer lattice constant, surface roughness, and wafer thickness, and the expected performance indicators include wavelength, power, and modulation rate. Obtaining the expected epitaxial property characteristics of the epitaxial layer includes: With the wafer attribute information as a constraint and wafer epitaxial growth as a guide, retrieve and obtain a sample data set. Among them, the sample data includes sample performance indicators and sample epitaxial property characteristics, and the sample epitaxial property characteristics include material composition, epitaxial layer thickness, and crystal surface profile. Use the sample data set to perform supervised training on a BP neural network to obtain an attribute analyzer. Parse the expected performance indicators through the attribute analyzer and output the expected epitaxial property characteristics. Among them, executing the epitaxial growth control of the wafer to be processed according to the epitaxial growth control parameters, multiple growth intervals, and multiple expected growth indicators includes: S1: Select a first growth interval and a second growth interval, and the second expected growth indicator of the second growth interval. Among them, the first growth interval is the first growth interval to start among the multiple growth intervals, and the second growth interval is the adjacent growth interval of the first growth interval. S2: According to the epitaxial growth control parameters, perform epitaxial growth of the first growth interval on the wafer to be processed. Among them, the epitaxial growth control parameters include growth temperature, reaction gas flow rate, and reaction gas ratio. S3: Use reflection high-energy electron diffraction fixed-point monitoring to obtain a crystal growth data set. Among them, the crystal growth data set includes an epitaxial layer thickness set and a crystal surface profile set. S4: Perform crystal growth prediction within a predetermined window according to the epitaxial layer thickness set and the crystal surface profile set to obtain the first predicted epitaxial property data at the end node of the first growth interval. S5: Calculate the deviation between the second expected growth indicator and the first predicted epitaxial property data to determine the second growth deviation. S6: With meeting the second growth deviation and the second growth interval as constraints, use the TACD simulation tool to perform reverse simulation to determine the second epitaxial growth control parameters. S7: Execute the epitaxial growth control of the second growth interval according to the second epitaxial growth control parameters. After executing the epitaxial growth control of the second growth interval according to the second epitaxial growth control parameters, iterate through S1 to S7 until the predetermined epitaxial growth cycle is completed. Among them, predicting crystal growth within a predetermined window according to the epitaxial layer thickness set and the crystal surface profile set includes: Performing thickness growth trend analysis according to the epitaxial layer thickness set to determine the thickness growth rate; Selecting the terminal epitaxial layer thickness from the epitaxial layer thickness set, and predicting crystal growth within the predetermined window according to the thickness growth rate and the terminal epitaxial layer thickness to obtain the first predicted epitaxial layer thickness; Performing profile change trend analysis according to the crystal surface profile set to determine the profile change characteristics; Selecting the terminal surface profile from the crystal surface profile set, and predicting crystal growth within the predetermined window according to the profile change characteristics and the terminal surface profile to obtain the first predicted crystal surface profile; Obtaining the first predicted epitaxial property data based on the first predicted epitaxial layer thickness and the first predicted crystal surface profile; Taking the sum of the monitoring data delay, the prediction data delay, and the parameter control delay as the predetermined window; Obtaining the first predicted epitaxial property data includes: Using the TACD simulation tool to perform growth simulation of the wafer to be processed according to the first growth interval and the epitaxial growth control parameters to obtain the first simulated epitaxial layer thickness, the first simulated crystal surface profile, and the first simulated material composition; Calculating the deviation between the first predicted epitaxial layer thickness and the first simulated epitaxial layer thickness to obtain the first predicted thickness difference; Performing deviation analysis on the first predicted crystal surface profile and the first simulated crystal surface profile to determine the first predicted profile deviation characteristics; Performing correlation correction on the first simulated material composition according to the first predicted thickness difference and the first predicted profile deviation characteristics to obtain the first predicted material composition; Setting the first predicted epitaxial layer thickness, the first predicted crystal surface profile, and the first predicted material composition as the first predicted epitaxial property data.

Images